jevans ⁂ @ thejevans @lemmy.ml

Air quality scientist and data engineer

---

Read more

Posts

14

Comments

636

Joined

3 yr. ago

I'll have to double check, but I came from a B450 board. It definitely allowed me to run my RAM at a higher XMP profile (4x 3200MHz), and it has way better IOMMU groups. Each PCIe device gets its own group, so they can all be passed to different VMs.

I have a similar setup. I just recently switched to the ASRock Phantom X570 for $100. It's a fantastic board at that price.

The same thing happened to me when I first tried to go there, but it's fine now.

https://pixelfed.social/p/thejevans/664709222708438068

EDIT:

Server:

• AMD 5900x
• 64GB RAM
• 2x10TB HDD
• RTX 3080
• LSI-9208i HBA
• 2x SFP+ NIC
• 2TB NVMe boot drive

Proxmox hypervisor:

• TrueNAS VM (HBA PCIe passthrough)
• HomeAssistant VM
• Debian 12 LXC as SSH entrypoint and Ansible controller
• Debian 12 VM with Ansible controlled docker containers
• Debian 12 VM (GPU PCIe passthrough) with Jellyfin and other services that use GPU
• Debian 12 VM for other docker stuff not yet controlled by Ansible and not needing GPU

Router: N6005 fanless mini PC, 2.5Gbit NICs, pfsense

Switch Mikrotik CRS 8-port 2.5Gbit, 2-port SFP+

I agree. Hence why I wondered why that would be an acceptable option compared to simply changing OPs posted requirements for far less cost and headache.

It's not a BIG deal. I self-host a ton of stuff. It's just a bigger deal than spending a bit more for phone storage for the vast majority of people.

The correlation is high.

If you're willing to spend all the money on setting up and running a server, why not just spend way less money to get more phone storage and use OSM?

It costs money to host something like that. You want low latency, real-time routing and tile-rendering? Even more money. Sure, it could be funded by donations or something like that, but I'm not holding my breath.

What you're asking for is fairly unrealistic. The only way this could work sustainably would be for something to exist where you host your own tile server and routing service and patch that into OSM. Otherwise, even if the app itself is open source, the backend will cost money to run and will be proprietary.

The reason that OSM is able to be fully open source is because you host the tiles on your phone and do the routing calculations locally.

If that's the case, then just set up a pipeline to pre-transcode your 4k content to 1080p, so your server doesn't have to handle that on the fly.

Alacritty is really nice and easy to configure, and isn't "tied" to any desktop environment, like Konsole is. Kitty is really cool for its implementation of image display. Foot is a Wayland-native alternative that is also really nice to use.

I use it to set and manage timers in the kitchen. It's not as good as Google, and the setup to get timers working is hacky, but it does the job and has fully replaced my Google Home when combined with a home assistant dashboard I have on my kitchen wall.

Are you me? <3

You took the words right out of my mouth.

It's not lost on me lol. The news site is also a .CO domain.

ICANN is certainly not perfect, but there is a difference between the automatic control that countries have over their ccTLD and the control they have through ICANN.

The article explains that, yes, they did plan to move...in April. The Taliban government did, in fact, shut them down ahead of that schedule.

The problem is who controls them. The government of each country can decide at any moment that they want to take control of their TLD and remove any sites that they don't like. It's just not good practice if you want your site to stick around.

Yes, and that would have happened months from now. The Taliban shut them down immediately.

Oh I'm fully aware haha, and the article is on a CO site. At the time I made my account, lemmy.ml seemed to be the best option, even with the poor TLD choice.